ReSplat

ReSplat: Learning Recurrent Gaussian Splats

Haofei Xu1,2     Daniel Barath1     Andreas Geiger2     Marc Pollefeys1,3
1ETH Zurich     2University of Tübingen, Tübingen AI Center     3Microsoft    
Paper

ReSplat enables recurrent Gaussian splats reconstruction in a feed-forward manner.

As initialization (iteration 0), we introduce a compact and expressive feed-forward model to produce 16x fewer Gaussians than pixel-aligned feed-forward models while still achieving better results.

ReSplat is able to benefit from additional test-time compute with more iterations.

Feed-forward novel view synthesis (512x960) on unseen scenes from 16 input views with ~500K Gaussians (16x fewer than pixel-aligned models).

Abstract

While feed-forward Gaussian splatting models provide computational efficiency and effectively handle sparse input settings, their performance is fundamentally limited by the reliance on a single forward pass during inference. We propose ReSplat, a feed-forward recurrent Gaussian splatting model that iteratively refines 3D Gaussians without explicitly computing gradients. Our key insight is that the Gaussian splatting rendering error serves as a rich feedback signal, guiding the recurrent network to learn effective Gaussian updates. This feedback signal naturally adapts to unseen data distributions at test time, enabling robust generalization. To initialize the recurrent process, we introduce a compact reconstruction model that operates in a 16x subsampled space, producing 16x fewer Gaussians than previous per-pixel Gaussian models. This substantially reduces computational overhead and allows for efficient Gaussian updates. Extensive experiments across varying of input views (2, 8, 16), resolutions (256x256 to 540x960), and datasets (DL3DV and RealEstate10K) demonstrate that our method achieves state-of-the-art performance while significantly reducing the number of Gaussians and improving the rendering speed.

Approach

Compact initialization with 16x fewer Gaussians than pixel-aligned models.
Recurrent Gaussian update by using the rendering error as a feedback.

State-of-the-art Performance

On RealEstate10K, ReSplat achieves performance similar to LVSM while offering 20x faster rendering speed thanks to its efficient 3DGS representation. On DL3DV, ReSplat surpasses previous methods by a clear margin, while using 4x-16x fewer Gaussians than pixel-aligned models.

Robust Generalization

ReSplat leverages the rendering error as a feedback signal, allowing it to adapt to the test data and thereby achieve better generalization to unseen scenarios, such as new datasets and resolutions.

High-Quality View Synthesis

ReSplat produces better novel view synthesis results than prior methods on DL3DV.

BibTeX

@article{xu2025resplat,
      title={ReSplat: Learning Recurrent Gaussian Splats},
      author={Xu, Haofei and Barath, Daniel and Geiger, Andreas and Pollefeys, Marc},
      journal={arXiv preprint arXiv:2510.08575},
      year={2025}
    }

awesome webpage template